Abstracts

Rationale: Responsive neurostimulation (RNS) is an FDA approved mode of neuromodulation for adults with drug-resistant focal epilepsy. However, data for use in pediatric patients is limited, especially when targeting specific seizure types. In this case series, we describe our initial experience with thalamic ictal EEG activity in pediatric patients with RNS implanted in the bilateral thalamic centromedian (CM) nuclei to detect and treat epileptic spasms (ES). We also report their short-term seizure outcomes.

Methods: In this single-center retrospective case series, we reviewed the intracranial thalamic data and clinical reports of 3 patients with pediatric-onset drug-resistant epilepsy (DRE) (age: 9, 15, and 15 years; 1 female) with RNS depth electrode implantation targeting bilateral CM and had identified ES as an active and disabling seizure type. Two patients underwent continuous video EEG (cvEEG) to characterize the thalamic ictal EEG pattern to guide RNS programming changes. The third patient underwent programming changes after reviewing RNS Patient Data Management System (PDMS) intracranial recordings (without cvEEG evaluation).

Results: When correlating clinical spasms, scalp EEG, and RNS thalamic activity, based on time-frequency analysis, we found that there is an increase in thalamic ictal gamma frequencies associated with slow waves that correlate with ES. These ictal changes in the CM occurred concurrently with scalp EEG changes of high amplitude slow wave followed by attenuation and clinical presentation of spasms (head drop, rapid flexion at the waist). Thus, RNS detection settings were changed to bandpass targeting higher gamma frequencies (28.85-125 Hz).
Adjustments to target gamma frequencies were made at a median 22 months (range 21-28) after initial RNS stimulation activation. Patients had clinic follow up a median 3.5 months (range 1-4) and most recent PDMS data (June 2024) was reviewed a median 4.5 months (range 4-8) after gamma detection setting changes were made.
Following adjustment to target gamma frequencies correlating with ES, magnet swiping events were accurately captured by the RNS detection. In the most recent follow up clinic notes, caretakers reported that spasms were increased, possibly reduced, and significantly reduced ( >50%) for each patient. According to the RNS PDMS database, all activity (daily total detections and therapies) from pre-gamma detection settings to most recent had a change in all events with a median decrease of 19% (range 93% decrease to 14% increase). Caretakers did not report any significant side effects.

Conclusions: The significance of these findings is that ictal gamma activity may serve as a thalamic EEG signature of ES. Simultaneous scalp EEG recordings can help identify such ictal segments to further adjust the RNS detection settings. A study with a larger sample size and longer follow-up is needed to determine the efficacy of bithalamic RNS in treating epileptic spasms.

Funding: No financial disclosures.